Current sampling methods and circuits typically use averaging circuits that can distort the baseline voltage reference of a sampled output signal. This is because sampled pulses are averaged in with the baseline voltage reference level. For example, negative going input pulses can, once averaged, result in the baseline voltage reference being lower than the desired baseline voltage reference level in the output signal. Alternatively, positive going input pulses can, once averaged, result in the baseline voltage reference being higher than the desired baseline voltage reference in the output signal. What is desired is a method and circuit for accurately reproducing the baseline voltage reference level in the sampled output signal without any voltage offsets despite the presence of input pulses.